A comparative study of soft-computing methodologies in identification of robotic manipulators

This paper investigates the identification of nonlinear systems by utilizin g soft-computing approaches. As the identification methods, feedforward neu ral network architecture (FNN), radial basis function neural networks (RBFN N), Runge-Kutta neural networks (RKNN) and adaptive neuro-fuzzy inference s ystems (ANFIS) based identification mechanisms are studied and their perfor mances are comparatively evaluated on a two degrees of freedom direct drive robotic manipulator. (C) 2000 Elsevier Science B.V. All rights reserved.